Research of Association Rules Algorithm Based on Matrix under Cloud Computing

Abstract

It occurs time and space performance bottlenecks when traditional association rules algorithms are used to big data mining. This paper proposes a parallel algorithm based on matrix under cloud computing to improve Apriori algorithm. The algorithm uses binary matrix to store transaction data, uses matrix "and" operation to replace the connection between itemsets and combines cloud computing technology to implement the parallel mining for frequent itemsets. Under different conditions, the simulation shows it improves the efficiency, solves the performance bottleneck problem and can be widely used in big data mining with strong scalability and stability.